| *** tosky has quit IRC | 00:00 | |
| openstackgerrit | Clark Boylan proposed zuul/nodepool master: Bump openshift dep https://review.opendev.org/c/zuul/nodepool/+/765873 | 00:02 |
|---|---|---|
| openstackgerrit | Clark Boylan proposed zuul/nodepool master: Bump openshift dep https://review.opendev.org/c/zuul/nodepool/+/765873 | 00:22 |
| *** imtiazc has joined #zuul | 00:27 | |
| *** ikhan has joined #zuul | 00:33 | |
| *** guillaumec has quit IRC | 00:52 | |
| *** guillaumec has joined #zuul | 00:52 | |
| *** Tahvok has quit IRC | 00:52 | |
| *** Tahvok has joined #zuul | 00:53 | |
| ianw | mordred: now with sorting, and icons :) | 00:56 |
| ianw | i've finished fiddling now. i'll just try copying guillaumec's bits for configuration, and that will be it | 00:56 |
| guillaumec | ianw, few things: do not set a link if "noop" or if "://" isn't part of the link, support for dark theme :) | 00:59 |
| ianw | ahh, cool will do. maybe i'll leave the dark theme for a future contributor | 01:03 |
| -openstackstatus- NOTICE: The Gerrit service on review.opendev.org is being restarted quickly to make heap memory and jgit config adjustments, downtime should be less than 5 minutes | 01:10 | |
| *** Goneri has quit IRC | 01:21 | |
| ianw | guillaumec: oh, that config isn't working the way i expect i don't think, now i read it more closely. we want to pick up a file on disk, rather than a config ref | 01:44 |
| guillaumec | ianw https://gerrit-review.googlesource.com/Documentation/dev-plugins.html#configuration | 01:48 |
| ianw | guillaumec: but can i get there from javascript? | 01:49 |
| guillaumec | ianw, I think so https://gerrit.googlesource.com/plugins/zuul-status/+/refs/heads/master/src/main/java/com/googlesource/gerrit/plugins/zuulstatus/GetConfig.java#41 switch getFromProjectConfigWithInheritance to getFromGerritConfig :) | 01:52 |
| guillaumec | getGlobalPluginConfig | 01:52 |
| corvus | ianw: do you want me to work on bootstrapping that into gerrit's gerrit? | 01:52 |
| ianw | corvus: the testing isn't done, i don't really fully understand the BUILD file and naming might change. if iterating on those things isn't a block, then yes :) | 01:54 |
| corvus | ianw: i think we only need to decide on a repo name; zuul-results ? | 01:54 |
| corvus | (it would be plugins/zuul-results actually) | 01:54 |
| ianw | i've been calling it zuul-summary-status but i'm not too fussed. zuul-summary-results, maybe? | 01:55 |
| ianw | zuul-results-summary? | 01:55 |
| corvus | ianw: wfm; i was only thinking 'results' vs 'status' since there's that live status plugin already | 01:56 |
| ianw | yeah, agree on that, it is reading the already produced results | 01:56 |
| corvus | ianw: i'll start on that tomorrow with zuul-results-summary or whatever else you decide is better and tell me to supercede that with overnight :) | 01:57 |
| ianw | ok, if i can get this config going, i think that's it | 01:57 |
| *** amotoki has quit IRC | 02:01 | |
| *** amotoki has joined #zuul | 02:02 | |
| ianw | guillaumec: re the getGlobalPluginConfig i was hoping to avoid the .java plugin bits. maybe i'm misunderstanding the bindings | 02:02 |
| *** hamalq has quit IRC | 02:12 | |
| *** rfolco has joined #zuul | 02:34 | |
| *** armstrongs has joined #zuul | 02:39 | |
| *** armstrongs has quit IRC | 02:59 | |
| *** bhavikdbavishi has joined #zuul | 03:02 | |
| *** bhavikdbavishi1 has joined #zuul | 03:06 | |
| *** bhavikdbavishi has quit IRC | 03:08 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 03:08 | |
| *** cloudnull has quit IRC | 03:18 | |
| *** cloudnull has joined #zuul | 03:19 | |
| *** rfolco has quit IRC | 03:24 | |
| *** Pilou has joined #zuul | 04:06 | |
| *** bhavikdbavishi has quit IRC | 04:17 | |
| *** bhavikdbavishi has joined #zuul | 04:37 | |
| *** vishalmanchanda has joined #zuul | 04:55 | |
| *** bhavikdbavishi1 has joined #zuul | 05:09 | |
| *** bhavikdbavishi has quit IRC | 05:09 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 05:09 | |
| *** bhavikdbavishi has quit IRC | 05:16 | |
| *** bhavikdbavishi has joined #zuul | 05:17 | |
| *** saneax has joined #zuul | 05:20 | |
| *** iamweswilson has quit IRC | 05:21 | |
| *** iamweswilson has joined #zuul | 05:21 | |
| *** evrardjp has quit IRC | 05:33 | |
| *** evrardjp has joined #zuul | 05:33 | |
| ianw | corvus: ok, everything cleaned up as best I can get it @ https://github.com/ianw/zuul-results-summary ... i think that is suitable for commit as the first attempt | 05:52 |
| *** zbr has quit IRC | 06:06 | |
| *** zenkuro has joined #zuul | 06:08 | |
| *** mach1na has joined #zuul | 06:58 | |
| *** mach1na has quit IRC | 07:03 | |
| *** bhavikdbavishi has quit IRC | 07:11 | |
| *** jfoufas1 has joined #zuul | 07:19 | |
| *** mach1na has joined #zuul | 07:24 | |
| *** bhavikdbavishi has joined #zuul | 07:36 | |
| openstackgerrit | Tobias Henkel proposed zuul/nodepool master: Bump openshift dep https://review.opendev.org/c/zuul/nodepool/+/765873 | 07:38 |
| *** mach1na has quit IRC | 07:41 | |
| *** mach1na has joined #zuul | 07:41 | |
| *** mach1na has quit IRC | 07:45 | |
| *** mach1na has joined #zuul | 08:02 | |
| *** tosky has joined #zuul | 08:02 | |
| *** hashar has joined #zuul | 08:06 | |
| *** jcapitao has joined #zuul | 08:06 | |
| *** rpittau|afk is now known as rpittau | 08:14 | |
| *** zenkuro has quit IRC | 08:28 | |
| *** saneax has quit IRC | 08:45 | |
| *** jpena|off is now known as jpena | 08:59 | |
| *** rfolco has joined #zuul | 09:00 | |
| avass | Did we have an announcement for https://review.opendev.org/c/zuul/zuul-jobs/+/742799 ? | 09:08 |
| ianw | avass: i never really got much comment, but if you'd like to drive it ... feel free :) | 09:43 |
| avass | ianw: Might do it later if I find time :) | 09:43 |
| avass | on another note I think it would be good if zuul could make deprecation notices from ansible visible, maybe that's another case for a warning build result | 09:45 |
| *** holser has quit IRC | 09:48 | |
| avass | but maybe that's another thing ansible hasn't standardized | 09:50 |
| *** holser has joined #zuul | 09:53 | |
| *** mach1na has quit IRC | 09:56 | |
| *** mach1na has joined #zuul | 09:59 | |
| *** nils has joined #zuul | 10:08 | |
| *** wuchunyang has joined #zuul | 10:17 | |
| *** ssbarnea has joined #zuul | 10:18 | |
| *** wuchunyang has quit IRC | 10:21 | |
| *** zbr has joined #zuul | 10:23 | |
| *** wuchunyang has joined #zuul | 10:35 | |
| *** ssbarnea has quit IRC | 10:36 | |
| *** wuchunyang has quit IRC | 10:39 | |
| *** bhavikdbavishi has quit IRC | 10:49 | |
| *** bhavikdbavishi has joined #zuul | 10:51 | |
| *** hashar is now known as hasharLunch | 10:56 | |
| *** saneax has joined #zuul | 11:00 | |
| openstackgerrit | Matthieu Huin proposed zuul/zuul-client master: API, CLI: Allow use of tenant-scoped API URLs https://review.opendev.org/c/zuul/zuul-client/+/765203 | 11:17 |
| openstackgerrit | Matthieu Huin proposed zuul/zuul-client master: encrypt: add --public-key argument https://review.opendev.org/c/zuul/zuul-client/+/765313 | 11:19 |
| *** zbr has quit IRC | 11:25 | |
| *** zbr has joined #zuul | 11:28 | |
| openstackgerrit | Matthieu Huin proposed zuul/zuul master: tools: Deprecate encrypt_secret.py, document zuul-client encrypt https://review.opendev.org/c/zuul/zuul/+/765316 | 11:32 |
| tobiash | zuul-maint: heads-up, cryptogrphy 3.3 has been released and broke lots of stuff here at various places | 11:33 |
| tobiash | one issue is https://github.com/pyca/cryptography/issues/5613 | 11:33 |
| *** zbr has quit IRC | 11:46 | |
| *** zbr has joined #zuul | 11:48 | |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: DNM: Test docker build due to cryptography 3.3 https://review.opendev.org/c/zuul/zuul/+/766205 | 11:53 |
| zbr | tobiash: not the only thing that broke widely in the last 24h :( | 11:54 |
| openstackgerrit | Tobias Henkel proposed zuul/nodepool master: DNM: Test docker build due to cryptography 3.3 https://review.opendev.org/c/zuul/nodepool/+/766206 | 11:54 |
| tobiash | zbr: what else? | 11:54 |
| zbr | i lost 3h this morning due to docker getting broken on centos8 due to 20.10 release, i had to move all containers to a fedora host. | 11:55 |
| tobiash | oh great | 11:56 |
| zbr | i wonder if our ensure-docker role still works. | 11:56 |
| openstackgerrit | Tobias Henkel proposed zuul/zuul-jobs master: DNM: Test ensure-docker due to new docker release https://review.opendev.org/c/zuul/zuul-jobs/+/766207 | 11:57 |
| tobiash | let's find out ^ | 11:57 |
| zbr | we were supposed to have a pipeline that tests them weekly but i have the impression that the pipeline was not enabled. | 11:58 |
| *** jcapitao is now known as jcapitao_lunch | 12:06 | |
| *** zenkuro has joined #zuul | 12:08 | |
| *** imtiazc has quit IRC | 12:09 | |
| *** mach1na has quit IRC | 12:13 | |
| *** jpena is now known as jpena|lunch | 12:35 | |
| *** wuchunyang has joined #zuul | 12:43 | |
| *** hasharLunch has quit IRC | 12:48 | |
| *** wuchunyang has quit IRC | 12:48 | |
| *** rlandy has joined #zuul | 13:01 | |
| *** mach1na has joined #zuul | 13:02 | |
| *** Goneri has joined #zuul | 13:24 | |
| *** zbr has quit IRC | 13:24 | |
| *** zbr has joined #zuul | 13:27 | |
| *** jpena|lunch is now known as jpena | 13:30 | |
| *** bhavikdbavishi has quit IRC | 13:30 | |
| zbr | is there a way in to trick zuul to skip most-jobs from running by adding something to a particular CR? either a comment, label, summary or a like a magic file? | 13:41 |
| zbr | imagine that you have 50 heavy jobs and for a particular urgent change that you know is safe, you want to run quickly, bypassing most of them (not all). | 13:43 |
| zbr | is there a way to configure it in a way that would you to do that? | 13:43 |
| mhu | zbr: you could gate it directly, and use "zuul promote" to put that change at the top of the queue | 13:44 |
| mhu | IIUC | 13:44 |
| *** zbr has quit IRC | 13:45 | |
| mhu | the use case being an urgent fix | 13:45 |
| *** zbr has joined #zuul | 13:47 | |
| *** jcapitao_lunch is now known as jcapitao | 13:48 | |
| *** saneax has quit IRC | 13:49 | |
| *** jfoufas1 has quit IRC | 13:55 | |
| *** rfolco is now known as rfolco|brb | 13:59 | |
| *** zbr has quit IRC | 14:03 | |
| *** ikhan has quit IRC | 14:04 | |
| *** zbr has joined #zuul | 14:04 | |
| *** ikhan has joined #zuul | 14:05 | |
| *** zenkuro has quit IRC | 14:05 | |
| *** johanssone has quit IRC | 14:23 | |
| *** fdegir0 has joined #zuul | 14:24 | |
| *** fdegir has quit IRC | 14:24 | |
| *** johanssone has joined #zuul | 14:26 | |
| *** noonedeadpunk has quit IRC | 14:30 | |
| *** noonedeadpunk_ has joined #zuul | 14:33 | |
| zbr | would anyone oppose the idea of starting to add type hints to zuul code? navigating current codebase, is bit confusing right now. | 14:37 |
| zbr | i see we already have mypy enabled, but very few types added, more like a demo. | 14:38 |
| corvus | zbr: it's already in progress, but type hints don't address confusion, they only address safety. if it's confusing, it needs documentation or better names. | 14:39 |
| zbr | i have to disagree here, type hinting is not used at runtime, so its existence does not make code safer. BUT adding and checking it forces us to write better code, and also documents expected times. | 14:40 |
| corvus | zbr: and yes, we're still experimenting. | 14:40 |
| corvus | zbr: it's used at type checking time, so it catches errors during unit test runs | 14:40 |
| zbr | i asked because I would be happy to help with that | 14:40 |
| corvus | that's what i mean by safety, it is intended to catch errors in code before running | 14:41 |
| zbr | i found them very useful, bit of PITA, but clearly useful. | 14:41 |
| corvus | however, through the entire experiment so far, i think it has caught one or two errors | 14:41 |
| corvus | so the jury is out as to whether it's worth the extra work | 14:41 |
| mordred | yah - it still takes some judgement - because sometimes the hinting stuff is still new/weird and the effort that has to be done to make mypy happy weird | 14:42 |
| corvus | it should not be relied on for documentation, it's better to document something if it's confusing | 14:42 |
| zbr | corvus: based on my last year experience adding them to molecule and ansible-lint: yep you do find lots of bugs. | 14:42 |
| mordred | s/weird/is weird/ | 14:42 |
| pabelanger | in a periodic pipeline, how could one do a job dependency across projects? For example, with container builds, we want to impose an order of jobs (via requires / provideds) and wondering how that would look like. My guess, is a central project where the jobs are defined, and setup the job graph | 14:42 |
| zbr | corvus: i mentioned that because I discovered how much i need to dig to discover that a parameter is supposed to be (working on adding support for "wip") | 14:43 |
| corvus | zbr: zuul has *extensive* unit tests which are usually sufficient to catch type bugs. i think the only one that mypy has caught was an inconsistent usage of a variable which had an equivalent string representation; in practice, it mostly didn't matter which one was used. technically an error, but not one that surfaced as a bug. all of that to say: the value may vary based on codebases. | 14:45 |
| corvus | zbr: yeah, you won't hear me arguing that some areas couldn't use more docstrings. :) | 14:45 |
| openstackgerrit | Felix Edel proposed zuul/zuul master: WIP Switch to ZooKeeper backed result event queues https://review.opendev.org/c/zuul/zuul/+/764344 | 14:47 |
| tristanC | after trying mypy --strict on many projects, i think the benefit really depends on the codebase, as mordred says it can be weird, and sometime mypy doesn't even work as expected | 14:49 |
| tristanC | which is unfortunate, because types can capture knowledge in a way understandable for both the computer and humans | 14:52 |
| mordred | yah. I'm a big fan in general of typing | 14:52 |
| mordred | but | 14:52 |
| zbr | tristanC: i guess it was not a big project. I failed to make --strict pass yet, just too much to do. | 14:52 |
| mordred | I'm a big fan of typing systems helping me more than I serve them - and of them working consistently | 14:53 |
| mordred | I think mypy and python typing is getting there - but is still young | 14:53 |
| zbr | mordred: corvus: as a start, please merge https://review.opendev.org/c/zuul/zuul/+/765557 to stop punishing py39 users which cannot even run tox -e linters right now. | 14:57 |
| corvus | zbr: we're not punishing anyone. if that was a careless use of phrase, please be more careful. if you really think we are, then please re-evaluate your assumptions of other people's motives. | 14:58 |
| zbr | fyi, if you run with py39: you will get other 6 syntax errors which were fixed in the newer versions. | 14:59 |
| zbr | it was joking, i am sure nobody has punishing plans around. | 14:59 |
| corvus | zbr: i don't understand the code change in 765557 | 14:59 |
| corvus | zbr: is that a mistake from mypy? | 15:00 |
| zbr | the code change is due to type change at runtime. | 15:00 |
| corvus | why is that a problem? this is python... :/ | 15:01 |
| corvus | being able to say "foo=1; foo='bar';" instead of "foo=1; baz='bar'; del foo;" is one of the great things about python :/ | 15:02 |
| zbr | is bad practice, anyway the real issue is: https://gist.github.com/ssbarnea/fffaec1521f0afed278d2cd89d4243f2 | 15:02 |
| zbr | you are not saving anything by reusing the variable name, is not cpu register ;) | 15:02 |
| corvus | it's not nothing | 15:02 |
| corvus | 'bad' is judgemental here. i think that code is fine. | 15:02 |
| corvus | this is exactly the sort of thing that makes me think mypy is a waste of time | 15:03 |
| zbr | yeah, same could be said about 90% of type issues: "works fine as is". The issue is that invites to mistakes when the code is touched again. | 15:03 |
| corvus | zbr: i understand what you're saying, but in practice, with all of the testing we have in zuul, it just hasn't been a very big issue | 15:04 |
| corvus | we've spent far more time dealing with mypy non-issues than we have fixing type bugs that slipped through because we didn't use mypy | 15:05 |
| corvus | and this example is exactly the sort of thing i'm talking about. here, mypy did not catch a bug, it just caused us to have an argument. | 15:05 |
| *** fdegir0 is now known as fdegir | 15:05 | |
| zbr | practical case: "change" parameter, is that a dictionary, a particular class instance or what? defining its type allow a developer to easily see how it can work with it, easily jump to the implementation and so on. | 15:07 |
| avass | for me adding type hinting just feels like the project realized it should have used a statically typed language :) | 15:07 |
| tristanC | zbr: perhaps add ` --allow-redefinition` | 15:07 |
| zbr | without type the editor has no close what it is, and user end-up grepping source code, and guessing. | 15:08 |
| zbr | using this approach we could also add a "#" as the start of the mypy in the end. | 15:09 |
| corvus | avass: yeah, my own opinion is i'd rather use a statically typed language designed for that from scratch, or use python as it was designed. i'm starting to feel mypy is worst of both worlds. | 15:09 |
| zbr | the fact that there is only one place where this is an issue is quite good, usually i found more issues likes this. | 15:09 |
| avass | corvus: I think you might want to take a look at the zuul-cache role whenever you got time, or at least the documentation since it tries to standardize how to pass artifacts between jobsf | 15:09 |
| corvus | avass: yes, quite likely. i have a few things on my plate this week; what kind of timeframe do you have in mind? | 15:10 |
| avass | corvus: yeah same here, my expeience with type hinting is that it's not worth it and it just solves a symptom | 15:10 |
| avass | corvus: there's no hurry yet since it would probably take a couple of weeks until we need it. and we could just put it in our internal zuul-jobs until it's in | 15:11 |
| zbr | i have the impression that the discussion goes into non-constructive direction. why to go into debases? | 15:11 |
| avass | corvus: mostly making sure you're aware :) | 15:12 |
| zbr | the reality is clear: "tox -e linters" fails under py39 due to syntax errors, this fixes a real issue. | 15:12 |
| zbr | probably tobiash should have included these errrors in the commit message, as they are unrelated to the changed source code. | 15:13 |
| avass | zbr: why does baseurl need to be changed to baseurl_str? | 15:13 |
| avass | oh that was tobiash change | 15:14 |
| zbr | because initiall baseurl is a list of strings, and latest is made a string. | 15:14 |
| corvus | zbr: i have no doubt that we will merge 765557 (i just reviewed it) to correct the error. however, the change provides feedback into our ongoing evaluation of mypy. i agree we've probably exhausted the productive part of the conversation for now. | 15:14 |
| avass | zbr: ah | 15:14 |
| zbr | i would say it: changing that file is a side-effect of his change, not the main fix. | 15:15 |
| corvus | sure, i understand the errors. but it's still an unecessary code change just to make mypy happy (like mordred was saying earlier) | 15:16 |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: Bump mypy to 0.790 https://review.opendev.org/c/zuul/zuul/+/765557 | 15:17 |
| tobiash | I just didn't know a quick way to allow redefinition, removed the code change ^ | 15:17 |
| corvus | tristanC: thanks! we learned something | 15:18 |
| avass | corvus: I'd argue that it's bad practice to reuse variables with different types to make the code easier to read. however in this case the function is so small so it shouldn't cause problems | 15:18 |
| zbr | well the current fix is not really a proper fix, mypy excludes should not done at cli parameters. they should be in config file | 15:18 |
| zbr | because this way the editor will not be able to indicate mypy errors. | 15:19 |
| zbr | anyway, i can do that in a follow-up. | 15:20 |
| *** Goneri has quit IRC | 15:23 | |
| *** Goneri has joined #zuul | 15:24 | |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: Bump mypy to 0.790 https://review.opendev.org/c/zuul/zuul/+/765557 | 15:29 |
| *** zbr has quit IRC | 15:29 | |
| *** zbr has joined #zuul | 15:30 | |
| openstackgerrit | Sorin Sbârnea proposed zuul/zuul master: Moved mypy config outside tox https://review.opendev.org/c/zuul/zuul/+/766245 | 15:35 |
| *** jfoufas1 has joined #zuul | 15:38 | |
| *** Shrews has joined #zuul | 15:38 | |
| Shrews | oh look, this channel is still here | 15:39 |
| *** bhavikdbavishi has joined #zuul | 15:39 | |
| corvus | quick everyone hide from Shrews! | 15:41 |
| corvus | Shrews: o/ :) | 15:41 |
| Shrews | lol o/ | 15:41 |
| *** bhavikdbavishi1 has joined #zuul | 15:42 | |
| *** bhavikdbavishi has quit IRC | 15:43 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 15:43 | |
| avass | Shrews: oh no this is a completely unrelated zuul project with the same users :) | 15:43 |
| Shrews | you all joined netflix then? | 15:44 |
| avass | yeah we pivoted | 15:45 |
| fungi | oh, is this netflix? i thought we were making the javascript test framework one | 15:45 |
| corvus | turns out we were all working for netflix all along! | 15:46 |
| *** zbr has quit IRC | 15:48 | |
| *** zbr has joined #zuul | 15:51 | |
| tobiash | o/ | 15:53 |
| pabelanger | more to my question about periodic pipelines, has anyone tested a dependent pipeline? I'm curious if that would work in the case of artifact builds | 15:54 |
| corvus | pabelanger: i think i understand the question, and i don't think it's been tried, but i don't think it's going to work. i think we would need new zuul features to have dependencies between enqueued refs in periodic pipelines. i think your other idea -- a central project that runs all the jobs for the different containers -- is the only way to do that today. | 15:58 |
| pabelanger | corvus: okay, that isn't an issue. I was curious if there was a fancier way of doing it | 15:59 |
| openstackgerrit | Sorin Sbârnea proposed zuul/zuul master: Updated importlib-resources dependency https://review.opendev.org/c/zuul/zuul/+/766253 | 16:02 |
| openstackgerrit | Matthieu Huin proposed zuul/zuul-client master: encrypt: add --public-key argument https://review.opendev.org/c/zuul/zuul-client/+/765313 | 16:02 |
| *** wuchunyang has joined #zuul | 16:02 | |
| *** wuchunyang has quit IRC | 16:07 | |
| zbr | corvus: tobiash: could we relook at https://review.opendev.org/c/zuul/zuul/+/763333 ? | 16:09 |
| zbr | apparently the UI changes the more people are involved the more we are sure nothing will happen. | 16:10 |
| corvus | zbr: that last remark is not how you convince people to work with you on changes. | 16:11 |
| * zbr nods | 16:11 | |
| corvus | zbr: i've added it to my list, but there are several things that are more pressing (mostly related to new gerrit), so it may be later this week or next before i have a chance | 16:12 |
| zbr | if the only reason is lack of time no problem, but working remotely sometimes you cannot make the difference between something just needing time or something that is undesired and nobody is able to sad it directly | 16:14 |
| corvus | zbr: i think people are interested in those changes, we just had a lot of urgent and surprising work recently | 16:15 |
| zbr | glad to hear that. and if there are bits that need more discussions I could easily remove them in order to allow it to progress. | 16:16 |
| zbr | indeed, gerrit 3.x related will likely keep most of us busy till next year | 16:17 |
| avass | oh is there a date for zuul 4 yet? | 16:20 |
| corvus | avass: no, but tobiash is working on cleaning up some patches; hopefully rsn | 16:23 |
| zbr | i would also want to add a py39 job to spot other issues related to py39, if nobody is against. | 16:23 |
| tobiash | corvus: I wanted to work today on it but then cryptography came... | 16:24 |
| *** zenkuro has joined #zuul | 16:25 | |
| tobiash | zbr: our policy is to test lowest and highest supported python so switching from py38 to py39 would be fine from my point of view | 16:27 |
| zbr | i would also include the shipped version in that list, which is py38 now. | 16:28 |
| zbr | i doubt we want to jump it to py39. | 16:29 |
| fungi | well, we ran into some problems with py39 recently which caused us to drop back to 38 for our images, right? | 16:29 |
| fungi | i guess we'd need to revisit whatever problems those were | 16:30 |
| zbr | but if we have a py37 we can skip that one and test only py36,py38,p39. usually the newest != production, is the "canary" one. | 16:30 |
| zbr | but if we do not run anything, we will not be able to spot issues. | 16:31 |
| *** mach1na has quit IRC | 16:31 | |
| zbr | maybe running py39 only at check would an ok compromise? | 16:31 |
| tobiash | currently we run just two versions in order to not overuse the opendev resources | 16:31 |
| corvus | if we can't run on 39 it's not worth testing; if we can run on 39 it's worth upgrading | 16:32 |
| zbr | that sounds like vicious cycle: you cannot know if something works until you test it,... | 16:33 |
| fungi | the word "test" is getting used in multiple ways here, i think | 16:34 |
| corvus | i think you misunderstand me -- as fungi said, we tried 3.9 and had problems and had to revert | 16:34 |
| corvus | yes, what fungi said | 16:34 |
| fungi | test whether it will work with 3.9, and if it does then set up automated testing to ensure it continues to work | 16:34 |
| zbr | did these problem produce a test that fails? | 16:34 |
| corvus | there's no point running unit tests on 3.9 if we know that there's some problem with the python interpreter that causes us to fail in production. if we think it will work, then we should add tests and do the upgrade. | 16:34 |
| fungi | zbr: i don't remember what the problems were, to be honest, i need to look back at old conversations to find out what the problem was unless someone happens to recall | 16:35 |
| corvus | i don't; tobiash or clarkb might recall what issues we had with 3.9? | 16:35 |
| zbr | but i was more of stability kind of issue if I understood correctly | 16:35 |
| fungi | it's also entirely likely that the problem was either a bug in python which got subsequently fixed, or a dep incompatibility which has since been resolved by whoever maintains it | 16:35 |
| tobiash | corvus: looking at the git log it looks like we had a problem with 3.8, reverted to 3.7, fixed the issue (755670) and switched to 3.8 again | 16:35 |
| clarkb | 3.9 has/had trouble with our console logger thing that runs on the remote hosts (but that is a different python than the services run under) | 16:35 |
| tobiash | afaik no one tried 3.9 yet | 16:35 |
| fungi | tobiash: oh! thanks, that's probably what i was remembering | 16:36 |
| clarkb | I believe the known issues with the remote console logger code have been addressed | 16:36 |
| clarkb | however, not in a release yet | 16:36 |
| corvus | ok, then it sounds like we can/should switch from 3.8 to 3.9 | 16:36 |
| fungi | either way, trying with 3.9 is worthwhile, and if it works then sure we should switch | 16:36 |
| fungi | i vaguely recall one problem was that older ansible versions couldn't work with newer python, and upgrading the base for the image we use in opendev would force us to drop ansible versions sooner than we intended | 16:38 |
| fungi | but that might have also been around 3.8 not 3.9 | 16:39 |
| corvus | i think we dropped some ansible versions recently? | 16:39 |
| zbr | i know for sure that even ansible 2.10 still has some problems with py39, minor so far. | 16:39 |
| corvus | but yeah, we should check that | 16:39 |
| fungi | i keep forgetting 3.9 has only been around for a month or two now | 16:39 |
| clarkb | zbr: I would avoid updating to 3.9 if ansible (a core component of zuul) is known to have problems with it | 16:39 |
| zbr | unrelated to python itself, but officially is not even listed in metadata. ansible-test fails to install some of its test requirements, but you can install them manually. | 16:40 |
| clarkb | in particular I think ansible issues tend to be exposed a bit less directly to users | 16:40 |
| fungi | since i start building and running alphas of the next python release as soon as they're tagged, by the time they release they're way older to me than they are to people consuming the release versions | 16:40 |
| tobiash | the ansible issue has been resolved by dropping support: https://review.opendev.org/c/zuul/zuul/+/727373 | 16:40 |
| tobiash | it was ansible 2.7 which had problems with python 2.8 | 16:41 |
| tobiash | 3.8 | 16:41 |
| fungi | aha, thanks for confirming | 16:41 |
| fungi | and yeah, i'm increasingly convinced we just haven't tried 3.9 yet | 16:41 |
| *** bhavikdbavishi has quit IRC | 16:45 | |
| avass | mnaser: the lgmt role is functional and only missing docs right? I feel like github not allowing the project owner to self approve is a bit of a limitation... | 16:46 |
| zbr | is there a way to run only unittest? the docs do not even mention how to run any tests. | 16:46 |
| *** rfolco|brb is now known as rfolco | 16:46 | |
| *** bhavikdbavishi has joined #zuul | 16:46 | |
| mnaser | avass: it's lacking one tiny thing, it doesn't read text inside review, so if you approve and type `/approve` in there, it won't work -- you'll have to do it inside a comment too | 16:46 |
| mnaser | but i think that's trivial to fix | 16:46 |
| mnaser | avass: but other than that, yes, it works 100% for us with no problems. and yes, no self approve is gross | 16:47 |
| fungi | zbr: can you be more precise with your question? which repo? which unit test(s)? | 16:47 |
| avass | mnaser: great, I'll try to get that set up than and maybe update the docs for it | 16:47 |
| fungi | zbr: also which docs are you expecting to explain that? | 16:48 |
| mnaser | avass: awesome! i think pabelanger also did some hacking so they might be able to chime in | 16:48 |
| zbr | "zuul" repo, "tox -va" reports something not quite clear. I happen to know that I can run "tox -e py3" if I want to run tests but there is no mention of what it would do. | 16:49 |
| pabelanger | mnaser: avass: yah, did some hacking 2 weeks ago, got something basic going. I think making zuul-lgtm python app, is the way to go. Or some other generic name. that lets us get the python bits of our zuul-jobs. | 16:49 |
| zbr | i would say that expecting to see the words "unit" "func" or "integration", would not be unresonable. | 16:50 |
| pabelanger | https://github.com/ansible-network/sandbox/pull/60 was my POC | 16:50 |
| pabelanger | running from local laptop | 16:50 |
| openstackgerrit | Merged zuul/zuul master: Bump mypy to 0.790 https://review.opendev.org/c/zuul/zuul/+/765557 | 16:50 |
| mhu | zuul-maint, if ^ got merged let's merge https://review.opendev.org/c/zuul/zuul-client/+/765553 as well | 16:52 |
| pabelanger | zbr: avass: I want to see about adding 'reviewer' approvel thing next | 16:52 |
| pabelanger | as github app | 16:52 |
| *** hamalq has joined #zuul | 16:53 | |
| mnaser | pabelanger: if you take that route, i guess it'll require installing 2 apps | 16:54 |
| *** hamalq_ has joined #zuul | 16:55 | |
| pabelanger | yah | 16:55 |
| pabelanger | that is the downside, I'm still scared of adding github app creds into zuul jobs. More a mental thing | 16:55 |
| pabelanger | for zuul ci system | 16:55 |
| *** hashar has joined #zuul | 16:56 | |
| *** hamalq has quit IRC | 16:59 | |
| openstackgerrit | Sorin Sbârnea proposed zuul/zuul master: Decouple mypy from tox https://review.opendev.org/c/zuul/zuul/+/766245 | 17:05 |
| *** zenkuro has quit IRC | 17:08 | |
| *** rpittau is now known as rpittau|afk | 17:12 | |
| fungi | zbr: neat, i had no idea that `tox -va` provided a list of testenvs, nor that it was possible to add descriptions | 17:12 |
| zbr | fungi: yep, very useful. sadly it does not have an option to always display them without -v | 17:14 |
| zbr | i wanted to make a PR to add it when a tty is present, and maybe even use some colors. | 17:14 |
| *** zenkuro has joined #zuul | 17:14 | |
| zbr | but due to WIP work on 4.0 rewrite it would be a waste of time | 17:15 |
| zbr | meanwhile I worked on on tox-ansible plugin which can now detect and run ansible tests from both: molecule and ansible-test. | 17:16 |
| zbr | i mention this here because I do see a use-case for zuul-jobs in particular. | 17:17 |
| fungi | zbr: playing around with the description parameter, it looks like you can add a default description in the general [testenv] which will get inherited by implicit envs like py3 (but in that case you'll want to make sure all explicit testenvs also have descriptions so they don't inherit that) | 17:24 |
| zbr | fungi: i was doing the same myself | 17:24 |
| zbr | i know how to use selector to avoid the issue you mentioned | 17:25 |
| zbr | let me make a CR and you can edit it after to correct my poor descriptions | 17:25 |
| *** jfoufas1 has quit IRC | 17:26 | |
| *** zbr has quit IRC | 17:27 | |
| *** sugaar4 has quit IRC | 17:29 | |
| *** zbr has joined #zuul | 17:29 | |
| avass | I created a simple service that tracks a git repo containing a tenant config and serves it on an http api to be used with tenant_config_script. If anyone wants it I could probably push that upstream | 17:34 |
| avass | now currently it only tracks a single file but the idea is that it's supposed to be able to merge yaml files containing a single tenant as well. | 17:35 |
| *** zbr has quit IRC | 17:47 | |
| *** zbr has joined #zuul | 17:49 | |
| *** zbr has quit IRC | 17:56 | |
| *** zbr has joined #zuul | 17:58 | |
| *** noonedeadpunk_ is now known as noonedeadpunk | 18:01 | |
| *** wuchunyang has joined #zuul | 18:04 | |
| *** wuchunyang has quit IRC | 18:08 | |
| -openstackstatus- NOTICE: The Gerrit service on review.opendev.org is currently responding slowly or timing out due to resource starvation, investigation is underway | 18:17 | |
| *** reiterative has quit IRC | 18:19 | |
| *** reiterative has joined #zuul | 18:19 | |
| *** jcapitao has quit IRC | 18:20 | |
| *** fbo is now known as fbo|lunch | 18:28 | |
| *** fbo|lunch is now known as fbo|off | 18:28 | |
| *** hashar has quit IRC | 18:38 | |
| *** hashar has joined #zuul | 18:38 | |
| *** nils has quit IRC | 18:45 | |
| *** jpena is now known as jpena|off | 18:46 | |
| *** bhavikdbavishi has quit IRC | 19:00 | |
| *** hashar has quit IRC | 20:03 | |
| *** ikhan has quit IRC | 20:54 | |
| *** vishalmanchanda has quit IRC | 20:54 | |
| *** ikhan has joined #zuul | 20:58 | |
| *** rfolco has quit IRC | 21:08 | |
| *** Goneri has quit IRC | 21:21 | |
| *** smyers_ has joined #zuul | 21:30 | |
| *** smyers has quit IRC | 21:33 | |
| *** smyers_ is now known as smyers | 21:33 | |
| *** zenkuro has quit IRC | 21:37 | |
| *** imtiazc_ has joined #zuul | 22:02 | |
| *** imtiazc_ is now known as imtiazc | 22:04 | |
| *** wuchunyang has joined #zuul | 22:05 | |
| *** wuchunyang has quit IRC | 22:09 | |
| openstackgerrit | Guillaume Chauvel proposed zuul/zuul master: Update tests/base.py to use proper git data https://review.opendev.org/c/zuul/zuul/+/742746 | 23:15 |
| openstackgerrit | Guillaume Chauvel proposed zuul/zuul master: Fix gerrit merge commit change with zuul configuration https://review.opendev.org/c/zuul/zuul/+/762886 | 23:15 |
| openstackgerrit | Guillaume Chauvel proposed zuul/zuul master: Improve merger getFilesChanges, Fix edge cases https://review.opendev.org/c/zuul/zuul/+/762887 | 23:15 |
| openstackgerrit | Guillaume Chauvel proposed zuul/zuul master: Tenant reconfiguration: Add tests https://review.opendev.org/c/zuul/zuul/+/742747 | 23:15 |
| openstackgerrit | Guillaume Chauvel proposed zuul/zuul master: Tenant reconfiguration: Reorg scheduler event process loop https://review.opendev.org/c/zuul/zuul/+/742748 | 23:15 |
| openstackgerrit | Guillaume Chauvel proposed zuul/zuul master: Tenant reconfiguration: Scheduler ref-updated create/delete https://review.opendev.org/c/zuul/zuul/+/739198 | 23:15 |
| openstackgerrit | Guillaume Chauvel proposed zuul/zuul master: Tenant reconfiguration: Allow ref-updated newrev+oldrev reconfiguration https://review.opendev.org/c/zuul/zuul/+/742749 | 23:15 |
| openstackgerrit | Guillaume Chauvel proposed zuul/zuul master: Tenant reconfiguration: Get ref-updated modified files https://review.opendev.org/c/zuul/zuul/+/739078 | 23:15 |
| *** bodgix has joined #zuul | 23:18 | |
| *** bodgix has quit IRC | 23:22 | |
| *** imtiazc has quit IRC | 23:22 | |
| *** bodgix has joined #zuul | 23:23 | |
| *** imtiazc has joined #zuul | 23:25 | |
| *** rlandy has quit IRC | 23:28 | |
| *** zbr has quit IRC | 23:46 | |
| *** zbr has joined #zuul | 23:48 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!